Measure for Measure. Semi-cylindrical illuminance: a semi-conceived measure?

This is a repository copy of Measure for Measure. Semi-cylindrical illuminance: a semi-conceived measure?.

White Rose Research Online URL for this paper: http://eprints.whiterose.ac.uk/112512/

Version: Accepted Version

Article:

Fotios, S. orcid.org/0000-0002-2410-7641 (2017) Measure for Measure. Semi-cylindrical illuminance: a semi-conceived measure? Lighting Journal , 82 (2). pp. 34-35. ISSN 0950-4559

eprints@whiterose.ac.uk https://eprints.whiterose.ac.uk/

Reuse

Unless indicated otherwise, fulltext items are protected by copyright with all rights reserved. The copyright exception in section 29 of the Copyright, Designs and Patents Act 1988 allows the making of a single copy solely for the purpose of non-commercial research or private study within the limits of fair dealing. The publisher or other rights-holder may allow further reproduction and re-use of this version - refer to the White Rose Research Online record for this item. Where records identify the publisher as the copyright holder, users can verify any specific terms of use on the publisher’s website.

Takedown

If you consider content in White Rose Research Online to be in breach of UK law, please notify us by

1

MEASURE FOR MEASURE

Semi-cylindrical illuminance: a semi-conceived measure?

Lighting Journal, February 2017; 82(2); 34-35

Semi-cylindrical illuminance has been promoted as a better measure for lighting design than vertical illuminance. Professor Steve Fotios is unconvinced

Steve Fotios is Professor of Lighting and Visual Perception within Sheffield University School of Architecture

Illuminance is a measure of the amount of light reaching a surface, more precisely defined

as the total luminous flux incident on a surface, per unit area. Road lighting standards for

subsidiary roads tend to focus on horizontal surface illuminance, light incident upon the road

surface. It may also be desirable to light vertical surfaces, such as faces and building

facades, in which case there may be a need to specify also vertical surfaces target

illuminances. Some texts suggest that semi-cylindrical illuminance is a better measure than

vertical illuminance, arguing this is because the face is not flat and light on sides of the face

also contribute to its visibility.

Semi-cylindrical illuminance is the averaged illuminance on the curved surface of an upright

semi-cylinder. It has been promoted as a better measure than vertical illuminance for

characterising the ability to evaluate faces apparently because it sounds like that should be

the case, and so is assumed to be despite the lack of factual evidence. For example, one

source states that: ‘semi cylindrical illuminance has a significant “side lighting” effect which

the vertical illuminance has none. Semi cylindrical illuminance brings out the roundness, the

three dimensionality of the human form’ [1]. See also guidance from IESNA [2], which states: ‘For a number of reasons pure vertical illuminance from whatever direction is not the

optimum parameter. The comparatively recent introduction of the concept of semi-cylindrical

illuminance has therefore been included in this guide’ but without revealing what those

reasons were nor how they are overcome by using semi-cylindrical illuminance.

Selecting the lighting class for a subsidiary road (BS 5489-1:2013, Tables A.5 and A.6) leads

to a specific P-class, the conditions for which are given in Table 3 of EN 13201-2:2015,

shown here as Table 1. This specifies a minimum semi-cylindrical illuminance in each class

as an ‘additional requirement if facial recognition is necessary’. In areas where there are

particular concerns about crime, a need for facial recognition, or where CCTV is present,

ES-2 series of lighting classes (Table 5 in BS EN 13201-2:2003). These were later replaced by the

SC lighting classes EN 13201-2:2015 but are otherwise identical, a series of nine classes of

semi-cylindrical illuminance ranging from 0.5 to 10 lux.

Lighting Class Average horizontal illuminance (lux) Minimum horizontal illuminance (lux)

Additional requirement if facial recognition is necessary.

Minimum maintained vertical illuminance (lux) Minimum semi-cylindrical illuminance (lux)

P1 15 3.0 5.0 5.0

P2 10 2.0 3.0 2.0

P3 7.5 1.5 2.5 1.5

P4 5.0 1.0 1.5 1.0

P5 3.0 0.6 1.0 0.6

P6 2.0 0.4 0.6 0.2

Table 1. P lighting classes for pedestrians and pedal cyclists [EN 13201-2:2015].

ORIGIN WITHIN ROAD LIGHTING

The concept of semi-cylindrical illuminance within road lighting design appears to

have originated in articles by Caminada and van Bommel [3] and Rombauts et al [4].

Subsequent works perpetuating semi-cylindrical illuminance tend to refer to these

two studies without questioning the basis of their proposals.

Rombauts et al investigated facial recognition and alleged a ‘very good correlation

between the facial recognition distance and the value of ESC.’ However, they did not

consider any other metrics of illuminance (horizontal, vertical, etc) and so are not

able to state whether semi-cylindrical illuminance was better or worse than these.

That is not sufficient reason to favour semi-cylindrical illuminance ahead of any other

measure: they would very likely have found equally good correlation if using vertical

illuminance but did not raise that question.

The investigation by Caminada and van Bommel did investigate different metrics but

their analysis of the test results was not sufficiently robust to support the concluded

3

measured during their facial recognition test: data were presented, however, for only

two of these, vertical and semi-cylindrical illuminances, but not the third,

hemispherical illuminance. They claimed that there was ‘no reliable correlation’

between face recognition distance and vertical illuminance, but ‘good correlation’

between semi-cylindrical illuminance and face recognition distance. This conclusion

is not convincing because the degree to which correlation exists should be

determined by statistical analysis and that was not reported in this work.

There is, however, evidence which rejects semi-cylindrical illuminance. Boyce et al

[5] found a high degree of correlation between the cylindrical and horizontal

illuminances derived from cubic illuminances measured at 25 outdoor locations,

concluding that either ‘could be equally well related to subjects’ opinions’ and chose

therefore to use horizontal illuminance.

Alferdinck et al [6] concluded following a face recognition experiment that

semi-cylindrical illuminance did not give a better prediction of the results than vertical

illuminance, but this report did not present a statistical analysis to support that

conclusion.

Simons et al [7] analysed the ‘overall impression’ of road lighting in 12 locations and

commented that “The results taken in total indicate that hemispherical and semi-cylindrical illuminance give slightly worse correlation with the appraisals than does

horizontal illuminance. On the basis of the appraisals, therefore, there appears to be

no advantage in adopting hemispherical or semi-cylindrical in preference to

horizontal illuminance.” Again, however, there are no reported results to validate the

claim.

SUMMARY

While road lighting standards tend to focus on horizontal surface illuminance, it may

also be desirable to see vertical surfaces such as faces and building facades. These

are likely to be illuminated anyway by road lighting, either by direct illumination from

lamps or indirect illumination from reflection from other surfaces. It was assumed in

BS5489-1:2003 that ‘The provision of lighting designed to meet the requirements of

4

illuminance at the height of the human face, ensuring a high possibility of

recognition’. The move toward lanterns of greater cut-off and more-precise optics

may have led to a need now to specify target illuminances for vertical surfaces rather

than simply assuming this is the case.

Some design guides recommend values of semi-cylindrical illuminance. There is no

credible evidence (yet found) showing an advantage to using semi-cylindrical rather

than vertical illuminance. Semi-cylindrical illuminance is a widely-repeated possibility

that has yet to be substantiated. Because semi-cylindrical illuminance is still a single

number quantifying a magnitude of light, this alone does not provide more

information about the distribution of light in a space than does vertical illuminance,

such as light reaching the sides of a face. Promoting semi-cylindrical illuminance

would require designers and installers to invest in new meters for which there would

be reluctance unless a clear advantage can be shown.

The conclusion therefore is that designers should consider vertical illuminance but

not semi-cylindrical illuminance. This does not mean that semi-cylindrical illuminance

is an irrelevant metric, but rather that there does not yet appear to be any clear and

robust evidence for its need, nor for the values that should be targeted.

References

1 Report on energy saving opportunities in streetlighting. June 1999.

http://genesisnow.com.au/244-Energy%20Saving%20Opportunities%20in%20Streetlighting%20-%20Rep8.pdf

2 Recommended Lighting for Walkways and Class 1 Bikeways. DG5:1994. Illuminating Engineering Society of North America.

3 Caminada JF, van Bommel WJM. New lighting criteria for residential areas. Journal of the Illuminating Engineering Society, 1984; 13(4); 350-358.

4 Rombauts, P., Vandewyngaerde, H. & Maggetto, G. Minimum semicylindrical illuminance and modelling in residential area lighting. Lighting Research and Technology, 1989; 21; 49-55.

5

6 Alferdinck JWAM, Hogervorst MA, Van Eijk AMJ, Kusmierczyk JT. Mesopic vision and public lighting – A literature review and a face recognition

experiment. TNO-DV C435: 2010. The Netherlands.